v0.10.13.x RC5 Testing

Status
Not open for further replies.

moli

Grizzled Member
Aug 5, 2014
3,261
1,837
1,183
I have Options settings 5000 tdrk / 8 rounds. For several hours only 1000 tdrk got anonymized, and there are two large amounts of 4096 that have not split up. My guessing is there're not enough compatible denominations on testnet right now. But it seems large amounts have a hard time to anonymize.
 

Tesquenure

Member
Mar 23, 2014
110
13
78
I have Options settings 5000 tdrk / 8 rounds. For several hours only 1000 tdrk got anonymized, and there are two large amounts of 4096 that have not split up. My guessing is there're not enough compatible denominations on testnet right now. But it seems large amounts have a hard time to anonymize.
settings 1900 tDRK / 8 rounds
 

moli

Grizzled Member
Aug 5, 2014
3,261
1,837
1,183
settings 1900 tDRK / 8 rounds
Hey, thanks. :) I've set my wallet the same. Let's see what we'll get, and I'm going to bed soon, night time here. I'll leave the computer on. :)
 
Last edited by a moderator:

kryptofoo

Member
Jul 21, 2014
114
36
78
Germany
My linux qt client v101307 is set to anon 1000 tDRK, 4 rounds. I have 1048 anonymized and have completed the 4 rounds. But the status bar shows 69%. When I mouseover the status bar is says inputs have an average of 4 of 4 rounds. Status bar bug?
 

qwizzie

Well-known Member
Aug 6, 2014
1,549
728
183
i'm testing it again on v.0.10.13.7 after having made a payment to myself of all my coins (19.999).

So far everything below input amount 1024 seems to three-client pair very fast,
i'm keeping an eye on the larger input amounts (1024, 2048, 4096).

I'm off to work now and see what happens when i'm back
 

Coins101

New Member
May 14, 2014
20
4
3
"Use 3 clients in denominations to reduce the chance of sybil attacks.
This might cause a little blockchain bloat, but I think it's worth it."

http://jira.darkcoin.qa/browse/DRK-76

Would it be possible to connect to peers, partial mix, disconnect and mix with another set of peers as part of going around the master node network?
 

Kai

Member
Apr 6, 2014
110
56
78
"Use 3 clients in denominations to reduce the chance of sybil attacks.
This might cause a little blockchain bloat, but I think it's worth it."

http://jira.darkcoin.qa/browse/DRK-76

Would it be possible to connect to peers, partial mix, disconnect and mix with another set of peers as part of going around the master node network?
Meaning that the peers use a random masternode for each round : if they were together in r1 (round 1) they may not be together in r2 etc.
 

Coins101

New Member
May 14, 2014
20
4
3
Meaning that the peers use a random masternode for each round : if they where together in r1 (round 1) they may not be together in r2 ?
Yes.

If I try to snoop, I have no way to keep track.

If there are three peers as well, my data would become gibberish. I would have to set-up a sustained effort over a very long time to piece together a jigsaw puzzle, that might not yield any results.
 
  • Like
Reactions: Kai

Kai

Member
Apr 6, 2014
110
56
78
Yes.

If I try to snoop, I have no way to keep track.

If there are three peers as well, my data would become gibberish. I would have to set-up a sustained effort over a very long time to piece together a jigsaw puzzle, that might not yield any results.
Somehow I prefer the randomness, it could be more elegant. :D
 

oblox

Well-known Member
Aug 6, 2014
1,032
537
183
Meaning that the peers use a random masternode for each round : if they were together in r1 (round 1) they may not be together in r2 etc.
Isn't this the way it already is?
 

oblox

Well-known Member
Aug 6, 2014
1,032
537
183
Maybe right now for convenience, depending on how many round we choose, we stick with the same peers.
There is far less people testing than what mainnet would constitute so it makes sense that you are often times getting paired with the same people round after round. It should still be randomized though (just lacking right now during testing since it isn't representative of the end-user base of mainnet).
 

Coins101

New Member
May 14, 2014
20
4
3
Isn't this the way it already is?
That's what I thought.

I wasn't 100% sure why Kristov raised it with such a high priority, because of the randomness principle I thought was in the system already. It just needs a large pool of users.
 

stonehedge

Well-known Member
Foundation Member
Jul 31, 2014
696
333
233
That's what I thought.

I wasn't 100% sure why Kristov raised it with such a high priority, because of the randomness principle I thought was in the system already. It just needs a large pool of users.
Kristov was clear to say that he was observing testnet and not mainnet. Odd that he didn't highlight this partial mitigation.
 

moli

Grizzled Member
Aug 5, 2014
3,261
1,837
1,183
I'm going to put this in here because it's related to what we've been testing.... Kristov was on Freenode channel #darkcoin-test last night... I got to talk to him a little bit, and said to him that what i've been wondering about is exactly what's described under "Lonely Denomination Weakness".

Then I asked "how does DS check compatibility? like if A has all these 12 denominations, but B needs 13 denoms, then one of B's denom will not be anonymized? or B's wallet will just hang in limbo, waiting for C to come along with that denom to match?" - Here's his answer:
"each darksend peer knows how many and which denominations it will create in the next round, and broadcast that to other possible peers to make sure everyone is "compatible", "lonely" meaning that if I'm going to output 1 or more denominations of 50 DRK, for example, I will only peer with other people looking to create 1 or more denominations of 50 DRK"

I think we can ask him questions and hopefully he will help us with answers on the thread he created: https://darkcointalk.org/threads/pa...ckchain-privacy-via-darksend.2315/#post-20233 .
 
Last edited by a moderator:
  • Like
Reactions: Light

qwizzie

Well-known Member
Aug 6, 2014
1,549
728
183
okay, just got back from work and noticed this is the second time (also found this out with previous update) that large input amounts (4096, 2048, 1024) are not getting split up !!

Edit : and i received a lot of payments to myself without further denominations

 
Last edited by a moderator:

moli

Grizzled Member
Aug 5, 2014
3,261
1,837
1,183
okay, just got back from work and noticed this is the second time (also found this out with previous update) that large input amounts (4096, 2048, 1024) are not getting split up !!

Edit : and i received a lot of payments to myself without further denominations
Yes, I think there's some issue with splitting up and pairing large amounts. Please post on JIRA for us. Thanks.
 

coingun

Active Member
Masternode Owner/Operator
Jul 8, 2014
489
402
133
masternode.io
its already raised there but does not seem to be getting much attention : http://jira.darkcoin.qa/browse/DRK-74

Have sent an email to Evan with debug.log and wallet.
I also attempted to denominate large amounts. 1000 / 8 rounds, 5,000/8 rounds, 10,000/8 rounds, 15,000/8 rounds. They all got to various levels of complete but none of them actually finished. Might just be taking time though cause it's a lot of coins to anon.
 

TaoOfSatoshi

Grizzled Member
Jul 15, 2014
2,765
2,616
1,183
Dash Nation
www.dashnation.com
its already raised there but does not seem to be getting much attention : http://jira.darkcoin.qa/browse/DRK-74

Have sent an email to Evan with debug.log and wallet.
Guys, I created this Jira issue, but I ran it again (19000/ 8 rounds) and it worked. Please comment on this issue and attach your debug.log and your wallet.dat so that Evan can take a look.

Anyone can create an account on Jira, then post info.

Tao
 

Minotaur

Well-known Member
Foundation Member
Apr 7, 2014
452
1,079
263
For 1900 DRK / 8 rounds, i have 1998.00000024 DRK Anonymized, completion tel 98%
I think this is because the amount 1900DRK works as a threshold and it really means Not less than 1900 DRK, so the system will automatically target a number that is above 1900DRK and that suits the algorithm, so you are at 98% of whatever the target number the system picked is, it is not perfect but pretty damn good for what it is. This is just my interpretation without actually knowing exactly how it works internally.
 

qwizzie

Well-known Member
Aug 6, 2014
1,549
728
183
May i ask what kind of attention you are expecting? The ticket was created yesterday with Evan giving feedback short after... What SLA did we miss for you?
i ment that i had a bit of trouble locating this high prio ticket in the activity stream on JIRA and once i located it i found out that Evan was waiting
for a reply back that was in my opinion taking a bit long. So i mentioned it on darkcointalk to get the issue going again.
Then two hours ago TaoOfSatoshi made a post asking ppl to submit what i just did .. providing Evan with wallet and debug.log files when ppl experience these large input amounts not splitting / pairing problem..

I will try it again with a new wallet .. seeing Tao got it working in the end.
 
Last edited by a moderator:

qwizzie

Well-known Member
Aug 6, 2014
1,549
728
183
Same here:
Status 91%
Anon'd 5094.00000035 DRK
Target 5000 / 8 rounds
I my case i have Total balance of 1999.117 tDRK

What's can be the explanation. "Anonymize all" option can be the solution
if Anonymized = Target or if Anonymized > Target --> set Status to 100% and stop Anonymization ?
something like that ? (just thinking outloud)
 

mbilker

Member
Aug 7, 2014
55
8
48
If anyone needs to test large volume transactions, ask me, I have 10000 tdrk anonymized.
 
Status
Not open for further replies.